Fat Cell Size, Insulin Sensitivity, and Inflammation in Obese Children

CLAUDIO MAFFEIS, MD, DAVIDE SILVAGNI, MD, RICCARDO BONADONNA, MD, ALESSANDRA GREZZANI, MD, CLAUDIA BANZATO, MD,
AND LUCIANO TAT, MD

Objective To assess the association between adiposity indexes (body mass index [BMI], fat mass, adipocyte size) and
circulating inflammation markers with known metabolic relevance or insulin sensitivity in overweight/obese children.
Study design Twenty-eight children (males/females: 13/15) with different degrees of overweight (BMI z-score: 1.64-3.1; fat
mass: 14.1-49.8 kg) were studied. BMI, body composition (dual-energy x-ray absorptiometry scanning), subcutaneous
adipocyte diameter (needle biopsy of subcutaneous abdominal fat), blood tumor necrosis factor and interleukin-6 concentrations and insulin sensitivity (frequently sampled intravenous glucose tolerance test) were assessed.
Results Adipocyte diameter, more than BMI and fat mass, was significantly associated with interleukin-6 (r 0.62, P <
.001) and tumor necrosis factor (r 0.61, P < .001). Moreover adipocyte size was associated with insulin sensitivity
(R2 0.15, F 4.69, P .04) independently from fat mass.
Conclusions Adipocyte size is a factor linked to both inflammation and insulin resistance in overweight/obese children. This
is similar to the findings in adults and lends support to the tenet that the earlier obesity ensues, the more severe the biologic
consequences may be. (J Pediatr 2007;151:647-52)
hildhood obesity conveys a morbidity risk which can lead to hypertension, glucose intolerance, type 2 diabetes,
dyslipidemia, other related disorders and, when obesity persists, premature death in adulthood.1 Insulin resistance/
hyperinsulinemia and a microinflammatory state are commonly encountered in both adult and childhood obesity and
are believed to play a major role in the pathogenesis of the Metabolic Syndrome. Importantly, inflammation may promote insulin
resistance, hyperglycemia and hyperlipidemia and eventually atherosclerosis in human beings.2 Several inflammatory cytokines
can interfere with insulin signaling and impair insulin action.3 In previous studies, serum concentrations of tumor necrosis
factor (TNF-), interleukin-6 (IL-6), C-reactive protein and other inflammation related molecules were higher in obese
children than in non-obese subjects.4,5
Adipose tissue results in several inflammatory mediators being released into the systemic circulation, through which they
can exert systemic effects. The cellular origin of these inflammatory mediators may be not only adipocytes, but also, and perhaps
primarily, macrophages, which infiltrate fat depots to an extent that is proportional to the degree of obesity.6,7
In adults, adipocyte size is associated with both insulin resistance and inflammation. Obese people with hypertrophic
subcutaneous abdominal adipocytes display higher insulin levels and are more often glucose intolerant than equally obese
individuals with smaller subcutaneous abdominal adipocytes. Moreover, adipocyte size is
an independent predictor of type 2 diabetes in adults.8,9 To the best of our knowledge, no
data are available on the relationship between fat cell size, insulin resistance/hyperinsulinemia, and inflammatory markers in obese children.
From the Department of Mother and
The aim of this study was to assess the association between adiposity, fat cell size,
Child, Biology-Genetics, Section of Pediatrics (C.M., D.S., A.G., C.B., L.T.) and the
circulating inflammatory markers (IL-6, TNF-) and insulin sensitivity in a group of
Department of Biomedical and Surgical Scioverweight and obese prepubertal children. To do so, children were divided in tertiles
ences, Section of Endocrinology (R.B.), Unieither by BMI z score, or by body fat mass, or by fat cell size. The rationale of this
versity of Verona, Verona, Italy.
Submitted for publication Sep 12, 2006; last
approach was that for most clinicians, the BMI z score is the only surrogate index available
revision received Mar 16, 2007; accepted
to assess adiposity. Body fat mass, as assessed by dual-energy x-ray absorptiometry, is a
Apr 23, 2007.
true measure of adiposity, but it is available only in a number of centers. Finally, as a third
Reprint requests: Claudio Maffeis, MD, Department of Mother and Child, Biology-Gelevel of analysis, fat cell size may be a biologic measure, which, albeit related to total fat
netics, Section of Pediatrics, University of
mass, may play an independent role in obesity-related insulin resistance and inflammation.
Verona, P.le L.A. Scuro, 10, 37134 Verona,

BMI
IL-6
IS

Body mass index

Interleukin-6
insulin sensitivity

ISBSA
TNF-

IS normalized per m2 of body surface area

Tumor necrosis factor

Italy. E-mail: claudio.maffeis@univr.it.

0022-3476/$ - see front matter
Copyright 2007 Mosby Inc. All rights
reserved.
10.1016/j.jpeds.2007.04.053

647

Table I. Physical characteristics, fat cell diameter, insulin sensitivity and inflammatory markers: Subjects
are divided into tertiles, on the basis of their BMI z-scores

Number
Sex (M/F)
Age (yr)
Tanner stage (P1/P2)
Weight (kg)
Height (cm)
BMI (kg/m2)
BMI z-score
FM (%)
FM (kg)
FFM (kg)
Adipocyte diameter (m)
IL-6 (pg/mL)
TNF- (pg/mL)
ISBSA (ml/min/m2 per pmol/L)
IS (ml/min per pmol/L)

I tertile

II tertile

III tertile

Total

9
4/5
10.1 (1.9)
6/3
49.9 (9.6)
143.9 (8.6)
23.8 (1.9)
1.9 (0.1)
43.1 (7.6)
21.6 (5.7)
28.3 (6.3)
68.1 (9.1)
16.7 (16.2)
1.7 (1.5)
0.42 (0.20)
0.61 (0.28)

10
5/5
10.5 (1.6)
7/3
59.0 (9.8)
146.8 (7.4)
27.2 (2.2)
2.4 (0.1)
46.7 (8.2)
27.9 (8.2)
31.1 (5.4)
86.6 (16.0)
45.9 (28.7)
4.5 (2.9)
0.35 (0.13)
0.57 (0.21)

9
4/5
9.1 (0.8)
6/3
57.4 (13.1)
140.7 (10.0)
28.6 (2.7)
2.7 (0.2)
54.9 (6.3)
31.9 (9.7)
25.4 (4.5)
86.7 (26.2)
41.7 (26.7)
3.3 (2.2)
0.41 (0.23)
0.61 (0.33)

28
13/5
9.9 (1.5)
19/9
55.6 (11.3)
143.9 (8.7)
26.5 (2.9)
2.3 (0.4)
48.2 (8.7)
27.2 (8.8)
28.4 (5.7)
80.6 (19.8)
35.2 (27.1)
3.2 (2.5)
0.39 (0.18)
0.60 (0.21)

P value

.14
.18
.27
.002
.001
.02
.03
.11
.03
.05
.07
.88
.92

Data are shown as mean (SD).

METHODS
Subjects
Children (13 males and 17 females) with different levels
of overweight participated in the study. Lean children were
not studied because our Ethical Committee does not allow
the inclusion of healthy children in this clinical research study.
Two children were excluded from analysis because of impaired glucose intolerance demonstrated by oral glucose tolerance testing. The physical characteristics of the 28 remaining children are shown in Table I. Children with a BMI above
the 85th percentile were defined as overweight, and children
with a BMI above the 95th percentile were defined as obese.
National BMI tables were used as reference.10 On the basis of
Tanner stages the children were divided into 3 groups: prepubertal boys with no pubic hair and gonadal stage I, girls
with no pubic hair stage and breast stage I; pubertal boys
with pubic hair or gonadal stage II and girls with pubic hair
or gonadal stage II; late-postpubertal boys with voice
change and girls with menarche.11 None of the children had
any overt disease other than obesity nor were on any medication. Informed consent was obtained from the children and
their parents before taking part in the study. The protocol was
in accordance with the 1975 Declaration of Helsinki, as
revised in 1983, and was approved by the Ethical Committee
of Verona City Hospital.
Experimental Design
The study was cross-sectional and was carried out on 2
separate days, during which the children were under medical
supervision. In the days preceding the tests, no attempt was
made to influence the usual diet of each child (who had free
access to food). For 2 days preceding the tests, they did not
engage in any strenuous physical activity. Each child arrived at
648

Maffeis et al

the Department of Pediatrics at 8:30 A.M. on the day of the

adipose tissue biopsy, after an overnight fast (10 to 12 hours).
After anthropometry and body composition measurements
were taken, the children, in a postabsorptive state, underwent
needle aspiration of abdominal subcutaneous adipose tissue 5
to 8 cm laterally to the navel. Local anesthesia was obtained
with 1% Xylocaine, and subcutaneous adipose tissue 200 to
400 mg was aspirated with a 15-gauge needle.
On the second test day, no later than 4 days after the
adipose tissue biopsy, cytokine levels were measured from a
blood sample, and insulin sensitivity was assessed by means of
a frequently sampled intravenous glucose tolerance test.12 The
children arrived at the Department of Pediatrics at 8:30 A.M.
after an overnight fast. A topical anesthetic was applied to
both arms, and a 23-g Teflon catheter was inserted into the
antecubital vein of each arm, one for intravenous glucose
infusion, the other for blood sampling. After an appropriate
rest to establish baseline, blood samples were collected at
time 10= to determine IL-6 and TNF- levels and at time
10= and 0= to determine baseline values of glucose and
insulin. At time 0=, an intravenous infusion of 25% dextrose
was started at a rate designed to deliver 12 g of glucose per m2
of body surface area over approximately 1 minute. The time
required to infuse the glucose load was recorded for each
subject. Blood samples were drawn from the opposite arm at
2, 4, 6, 8, 10, 15, 20, 25, 30, 40, 60, 80, 100, 120, 140, 160,
180, 200, and 220 minutes to determine glucose and insulin
concentrations. Blood samples were also collected to measure
C-peptide for a companion experiment.

Anthropometry and Body Composition

Height and weight were measured in postabsorptive
condition and with an empty bladder. Height was measured
to the nearest 0.5 cm on a standardized height board. Weight
The Journal of Pediatrics December 2007

was determined to the nearest 0.1 kg on a standard physicians

beam scale, with the subject dressed only in light underwear
and no shoes. The BMI was calculated as weight (kilograms)
divided by height squared (meters squared). Complete data
were obtained from all of the children at baseline. BMI
z-scores were calculated with the LMS method and national
reference values of BMI and LMS coefficients.10 The children also underwent total body dual-energy X-ray absorptiometry to assess body composition with a DPX-L densitometer (Lunar Corp., Madison, WI). The subjects were scanned
in light clothing while lying flat on their backs. On the day of
each test, the DPX-L was calibrated according to the manufacturers instructions. Body fat mass was obtained by multiplying the percentage of body fat by body weight.

Adipose Tissue Analysis

Fat cell size was determined by light microscopy on
whole mounts of unfixed tissue immediately after biopsy. The
adipocyte cell diameter was measured using a graded ocular
scale by two independent expert operators on the same samples. Reproducibility of the measure was quantified replicating the measure 3 times (range of CV: 1.5%-4.5%). For the
statistical analysis, an average of 100 cells from each sample
was used.
Serum Concentrations of Inflammatory Markers
Concentrations of cytokines in serum were assayed with
an enzyme-linked immunosorbent assay (ELISA) technique
(High sensitivity human enzyme-linked immunosorbent assay set by ImmunoTools for IL-6 and Sigma, St. Louis, MO,
batch no. CKH-200A for TNF-).
Intravenous Glucose Tolerance Test
Analysis of the insulin and glucose curves during the
intravenous glucose tolerance test followed the general strategy proposed by Bergman and Cobelli,13,14 with slight modifications, as previously described.15 Values were estimated by
implementing this minimal model of glucose metabolism in
the SAAM II 1.1.2 software (SAAM Institute, Seattle, WA).
Numerical values of the unknown variables were estimated by
use of nonlinear least squares. Weights were chosen optimally, that is, equal to the inverse of the variance of the
measurement errors, which were assumed to be additive,
uncorrelated, with zero mean, and a constant coefficient of
variation (CV 2.5%).
The main outputs of this model are as follow:
1. Insulin sensitivity (SI): expressed as the increase in glucose
clearance at steady state, elicited by an increase in 1
pmol/L of insulin (units: mL min1 per pmol/L)
2. Glucose effectiveness (SG): expressed as an insulin-independent glucose clearance (units: mL min1)
Both SI and SG were normalized per m2 of body
surface area. The coefficients of variation of SI and SG were
Fat Cell Size, Insulin Sensitivity, and Inflammation in Obese Children

75.3 and 29.2%, respectively. The acute insulin response

(AIR) was also calculated in the first 10 minutes after glucose
infusion, as the area of serum insulin concentration above the
baseline values.16

Statistical Analysis
All results are shown as mean (SD). The KruskalWallis test was used to compare anthropometric characteristics, circulating inflammatory markers and insulin sensitivity
across tertiles of adiposity and of fat cell size. Post hoc analysis
by Tukeys test was also performed for between-group comparisons.
A zero-order correlation analysis was performed to assess the unadjusted association between physical characteristics, circulating inflammatory markers, insulin sensitivity, and
fat cell diameter. A multiple linear regression analysis with a
backward procedure was done with insulin sensitivity as the
dependent variable and fat mass (kg) and subcutaneous fat cell
diameter as independent variables.
A probability level of P .05 was used to indicate
statistical significance. The SPSS 13.0 (SPSS, Inc., Chicago,
IL) package for personal computers was used for all statistical
analyses.

RESULTS
Table I shows the physical characteristics and body
composition of the subjects divided into tertiles of BMI
z-score. Age, weight, and height were not statistically
different among the 3 groups. Fat mass (expressed both in
kilograms and as a percentage of body weight) but not
fat-free mass, was statistically different across the tertiles.
Subcutaneous adipocyte diameter was significantly different among tertiles. Neither circulating cytokines nor insulin sensitivity was significantly different across tertiles of
BMI z score.
Table II shows the physical characteristics and body
composition of the subjects divided into tertiles of fat mass
(kg). Age was not statistically different among the 3 groups.
However, height, weight, BMI, BMI z-scores, and fat mass
(expressed both in kg and as a percentage of body weight)
were all significantly different among the groups. Fat-free
mass was not statistically different across tertiles. Circulating
cytokines, insulin sensitivity, and subcutaneous adipocyte diameters were not significantly different among the groups.
Table III shows the physical characteristics and body
composition of the study subjects divided into tertiles of fat
cell diameter. Age, height, weight, BMI, BMI z-scores, fat
mass (expressed both in kg and as a percentage of body
weight), and fat-free mass were not statistically different
among the groups.
Circulating cytokines, but not insulin sensitivity, were
significantly different among the groups. In particular, IL-6
was significantly higher in the third tertile than in the first.
649

Table II. Physical characteristics, fat cell diameter, insulin sensitivity and inflammatory markers: Children
are divided into tertiles on the basis of their fat mass (kg)

Number
Sex (M/F)
Age (yr)
Tanner stage (P1/P2)
Weight (kg)
Height (cm)
BMI (kg/m2)
BMI z-score
FM (%)
FM (kg)
FFM (kg)
Adipocyte diameter (m)
IL-6 (pg/mL)
TNF- (pg/mL)
ISBSA (mL/min/m2 per pmol/L)
IS (ml/min per pmol/L)

I tertile

II tertile

III tertile

Total

9
4/5
8.9 (0.9)
7/2
44.3 (5.0)
136.4 (6.2)
23.7 (1.6)
2.1 (0.3)
40.7 (5.8)
17.9 (2.2)
26.4 (4.6)
72.5 (13.1)
26.0 (28.2)
2.3 (2.3)
0.44 (0.22)
0.60 (0.29)

10
5/5
10.4 (1.7)
6/4
56.3 (6.6)
147.4 (6.9)
25.8 (1.2)
2.2 (0.3)
47.3 (7.2)
26.3 (1.7)
30.0 (7.3)
82.9 (20.8)
41.1 (27.3)
3.9 (2.8)
0.41 (0.14)
0.65 (0.23)

9
4/5
10.3 (1.6)
6/3
66.0 (9.5)
147.6 (8.4)
30.1 (1.4)
2.7 (0.2)
56.7 (4.2)
37.4 (6.5)
28.6 (4.6)
86.3 (23.3)
37.7 (26.6)
3.3 (2.3)
0.31 (0.17)
0.53 (0.27)

28
13/5
9.9 (1.5)
19/9
55.6 (11.3)
143.9 (8.7)
26.5 (2.9)
2.3 (0.4)
48.2 (8.7)
27.2 (8.8)
28.4 (5.7)
80.6 (19.8)
35.2 (27.1)
3.2 (2.5)
0.39 (0.18)
0.60 (0.26)

P value

.095
.001
.007
.001
.002
.001
.001
.5
.288
.436
.32
.382
.61

Data are shown as mean (SD).

Table III. Physical characteristics, fat cell diameter, insulin sensitivity and inflammatory markers: Subjects
are divided into tertiles, on the basis of their adipocyte size

Number
Sex (M/F)
Age (yr)
Tanner stage (P1/P2)
Weight (kg)
Height (cm)
BMI (kg/m2)
BMI z-score
FM (%)
FM (kg)
FFM (kg)
Adipocyte diameter (m)
IL-6 (pg/mL)
TNF- (pg/mL)
ISBSA (mL/min/m2 per pmol/L)
IS (mL/min per pmol/L)

I tertile

II tertile

III tertile

Total

9
4/5
10.1 (1.4)
6/3
54.8 (10.1)
145.1 (7.6)
25.8 (3.0)
2.1 (0.3)
46.7 (8.9)
25.9 (8.3)
28.9 (5.4)
61.5 (5.1)
13.8 (17.2)
1.1 (1.5)
0.42 (0.20)
0.65 (0.30)

10
5/5
10.0 (2.1)
6/4
52.8 (11.9)
141.1 (11.2)
26.2 (2.9)
2.3 (0.4)
47.5 (8.7)
25.2 (7.4)
27.6 (7.4)
78.1 (3.9)
37.6 (25.5)
3.4 (1.7)
0.44 (0.17)
0.66 (0.22)

9
4/5
9.6 (0.9)
7/2
59.4 (11.7)
145.9 (6.7)
27.7 (2.9)
2.5 (0.3)
50.5 (9.0)
30.6 (10.6)
28.8 (4.3)
102.7 (17.6)
53.9 (23.2)
5.1 (2.5)
0.3 (0.17)
0.47 (0.24)

28
13/5
9.9 (1.5)
19/9
55.6 (11.3)
143.9 (8.7)
26.5 (2.9)
2.3 (0.4)
48.2 (8.7)
27.2 (8.8)
28.4 (5.7)
80.6 (19.8)
35.2 (27.1)
3.2 (2.5)
0.39 (0.18)
0.60 (0.26)

P value

.811
.59
.38
.39
.15
.67
.54
.64
.001
.004
.003
.15
.20

Data are shown as mean (SD).

TNF-, was significantly higher in the third than in the

second or in the first tertile.

Correlation Analysis
BMI z-score and fat mass (r 0.64; P .001), BMI
z-score and subcutaneous adipocyte size (r 0.50; P .01),
as well as fat mass and subcutaneous adipocyte size (r 0.38;
P .05) were significantly correlated. Both fat and adipocyte
size were inversely related (P .05) to insulin sensitivity
expressed as an absolute value (respectively r 0.34 and
r 0.46) and expressed as ISBSA (respectively r 0.38
and r 0.39). No association between BMI z-score and
650

Maffeis et al

insulin sensitivity was found. A multiple regression model,

with a backward procedure, showed that insulin sensitivity,
expressed as absolute value and as ISBSA, was significantly
predicted (respectively, R2 0.21, F 6.46, P .02 and
R2 0.15, F 4.69, P .04) by fat cell size, whereas fat
mass was excluded from the model. Fat cell size was significantly correlated also with AIR (r 0.50, P .012). However, when this correlation was adjusted for insulin sensitivity,
it was no longer statistically significant (P .07).
We found a positive correlation between either IL-6
(r 0.62; P .001) or TNF- (r 0.61; P .001) and
adipocyte diameter. IL-6 and TNF- were strongly associThe Journal of Pediatrics December 2007

ated (r 0.899; P .001). No association was found between

IL-6 or TNF- and fat mass or BMI z score. Finally, no
association was found between insulin sensitivity and inflammatory cytokines.

DISCUSSION
In this study, conducted in overweight and obese children, we explored the relationships of 3 different variables
related to adiposity, that is, BMI z score, body fat mass, and
fat cell size, with circulating inflammatory markers and insulin sensitivity. The latter ones are believed to reflect key
pathogenic mechanisms linking obesity to the traits of the
metabolic syndrome and to atherosclerosis. We reasoned that
they might be differently related to different facets of adiposity. In our study, neither BMI z score nor body fat mass were
associated with strong gradients in either inflammatory markers or insulin sensitivity. An inverse correlation between total
fat mass and insulin sensitivity was present, but it was weak.
In contrast, subcutaneous fat cell diameter was strongly related to inflammatory markers in a graded fashion. A weak
inverse correlation was also found between fat cell size and
insulin sensitivity. Accordingly, in a group of lean Pima
Indian children, who are at high risk for development of
obesity and diabetes, an association between abdominal adipocyte size and plasma glucose and insulin concentrations was
reported.17 However, at variance with the inflammatory
markers, the relationship between fat cell size and insulin
sensitivity was not graded, in that it showed a decline in
insulin sensitivity only in the top tertile of fat cell size.
Moreover, the positive correlation we observed between AIR
and fat cell diameter was no longer statistically significant
when adjusted for insulin sensitivity. Thus this correlation is
likely to be a consequence of the homeostatic inverse relationship linking insulin sensitivity to AIR.
Thus, of the 3 indexes related to adiposity, fat cell
diameter was the only one related to both inflammation and
insulin resistance, with a tighter relationship to the former
than to the latter. Increased adipocyte volume may be involved in the production and release of inflammatory signals
from adipose tissue, as a consequence of impaired fat cell
function or cell damage. The results of this study are an
extension of what we previously found in an ultramicroscopic
study of adipose tissue conducted in a sample of children
independent from this one, which showed areas of inflammation clearly detectable in fat tissue. The elementary lesion is
characterized by a microgranuloma with macrophage infiltration and lipodegenerative aspects.18 Now we know that fat
cell size, more than fat mass per se, and circulating indexes of
inflammation are associated in children, in close resemblance
to the pattern found in obesity in adult individuals.
Our data are consistent with the hypothesis that one
key event in the natural history of overweight/obesity is the
presence of large adipocytes, indicating that the bodys capability to store fat may be overwhelmed. The capability to
accommodate fatty substrate stores presumably depends on
both genetic and nongenetic factors regulating the number of
Fat Cell Size, Insulin Sensitivity, and Inflammation in Obese Children

adipocytes or adipocyte precursors, their capability to proliferate, calorie/fat intake, and energy expenditure. Once the fat
cell becomes hypertrophic, a cellular program is started, aiming to restrain further volumetric growth. Increased expression of cytokines and reduced synthesis of adiponectin lead to
cellular insulin resistance, in the presence of which further
growth of adipocytes is hampered severely. Hence, fatty substrates are redirected to other tissues, where the undesirable
effects of lipotoxicity occur, and ectopic fat deposition is
favored.
In this scenario, fat cell size is a closer indicator of
lipotoxicity than BMI or body fat mass. Indeed, subcutaneous
adipocyte diameter, but not BMI nor fat mass, is an independent predictor of type 2 diabetes, even after accounting for
insulin resistance and for beta-cell function.
In our data, fat cell diameter was more closely related to
inflammation markers than to insulin resistance. This finding
may have at least 2 explanations: (1) overexpression of TNF-
and IL-6 occurs in adipocytes and is strictly related to fat cell
size, whereas whole body insulin resistance is primarily determined by skeletal muscle and liver; and (2) we studied only
overweight/obese children, but, at least in adults, the fall in
insulin sensitivity is greater in the passage from lean to overweight than in that from overweight to overtly obese individuals. The same reasons may explain the lack of correlation
between inflammatory markers and insulin sensitivity in our
study.
Our data do not deny that BMI z score or fat mass are
related to both inflammatory markers and insulin resistance.
Rather, they emphasize that subcutaneous adipocyte diameter
is an adiposity measure more closely related to inflammation
and insulin resistance than the other 2, a finding which is in
agreement with the known biological relationships between
fat and metabolic diseases in adults. It may be hypothesized
that, because both BMI and fat mass increase across the
tertiles of fat cell diameter, they bear a relationship to inflammation and insulin resistance at least in part because they both
are surrogate indexes of adipocyte size.

ACKNOWLEDGMENT
This study received a partial sponsorship from the Ministry of
University and Research, Research Program of Remarkable National Interest (PRIN), 2006, protocol No. 67105, area No. 06.

REFERENCES
1.
Weiss R, Dziura J, Burgert TS, Tamborlane WV, Taksali SE, Yeckel CW, et al.
Obesity and the metabolic syndrome in children and adolescents. N Engl J Med.
2004;350:2362-74.
2.
Wellen KE, Hotamisligil GS. Inflammation, stress and diabetes. J Clin Invest
2005;115:1111-9.
3.
Dandona P, Aljada A, Chauduri A, Mohanty P, Garg R. Metabolic syndrome. A
comprehensive perspective based on interaction between obesity, diabetes and inflammation. Circulation 2005;111:1448-54.
4.
Aygun AD, Gungor S, Ustundag B, Gurgoze MK, Sen Y. Proinflammatory
cytokines and leptin are increased in serum of prepubertal obese children. 2005. Mediators Inflamm 2005;3:180-3.
5.
Reinher T, Stoffel-Wagner B, Roth CL, Andler W. High-sensitive C-reactive
protein, tumor necrosis alpha and cardiovascular risk factors before and after weight loss
in obese children. Metabolism 2005;54:1155-61.

651

6.
Lee YH, Nair S, Rousseau E, Allison DB, Page GP, Tataranni PA, et al.
Microarray profiling of isolated abdominal subcutaneous adipocytes from obese vs
non-obese Pima Indians: increase expression of inflammation-related genes. Diabetologia 2005;48:1776-83.
7.
Weisberg SP, McCann D, Desai M, Rosembaum M, Leibel RL, Ferrante AW.
Obesity is associated with macrophage accumulation in adipose tissue. J Clin Invest
2003;112:1796-808.
8.
Weyer C, Foley JE, Bogardus C, Tataranni PA, Pratley RE. Enlarged subcutaneous abdominal adipocyte size but not obesity itself, predicts Type II diabetes independent of insulin resistance. Diabetologia 2000;43:1498-506.
9.
Winkler G, Kiss S, Keszthelyi L, Sapi Z, Ory I, Salamon F, et al. Expression of
tumor necrosis (TNF)-alpha protein in the subcutaneous and visceral adipose
tissue in correlation with adipocyte cell volume, serum TNF-alpha soluble serum
TNF-receptor-2 concentrations and C-peptide level. Eur J Endocrinol
2003;149:129-35.
10. Luciano A, Bressan F, Zoppi G. Body mass index reference curves for children
aged 3-19 years from Verona, Italy. Eur J Clin Nutr 1997;51:6-10.
11. Tanner JM, Whitehouse RH. Clinical longitudinal standards from birth to
maturity for height, weight, velocity and stages of puberty. Arch Dis Child
1976;51:170-9.

652

Maffeis et al

12. Toffolo G, De Grandi F, Cobelli C. Estimation of -cell sensitivity from intravenous glucose tolerance test C-peptide data. Knowledge of the kinetics avoids errors in
modelling the secretion. Diabetes 1995;44:845-54.
13. Beard JC, Bergman RN, Ward WK, Porte D Jr. The insulin sensitivity index in
non diabetic men. Correlation between clamp-derived and IVGTT-derived values.
Diabetes 1986;35:362-9.
14. Gower BA, Nagy TR, Goran MI. Visceral fat, insulin sensitivity, and lipids in
prepubertal children. Diabetes 1999;48:1515-21. Erratum in: Diabetes 2001;50:477-78.
15. Maffeis C, Bonadonna RC, Consolaro A, Vettor R, Banzato C, Silvagni D, et al.
Ghrelin, insulin sensitivity and postprandial glucose disposal in overweight and obese
children. Eur J Endocrinol 2006;154:61-8.
16. Goran MI, Bergman RN, Cruz ML, Watanabe R. Insulin resistance and associated compensatory responses in African-American and Hispanic children. Diabetes
Care 2002;25:2184-90.
17. Abbott WG, Foley JE. Comparison of body composition, adipocyte size, and
glucose and insulin concentrations in Pima Indian and Caucasian children. Metabolism
1987;36:576-9.
18. Sbarbati A, Osculati F, Silvagni D, Benati D, Galie M, Camoglio FS, et al. Obesity
and inflammation: evidence for an elementary lesion. Pediatrics 2006;117:220-23.

The Journal of Pediatrics December 2007